OpenAI Decisions API vs Jev

Fact table

When to pick which

Pick a Jev-class decision endpoint when you need the pattern in production now: typed questions, per-option probabilities to threshold, and a callable API with no waitlist. Input is text and JSON state.

Watch OpenAI's Decisions API for when it opens broadly: the announcement promises image input and the GPT-6 Luna engine, but pricing, SDK shape, and rate limits are unpublished, so any migration is a plan, not a port.

Where this site fits

This site serves decisions-1 through its own endpoint — state plus 1 to 6 typed questions in, one answer per question with a probability for every option. New visitors get 2 free trial decisions; the call below is runnable from the playground.

curl -X POST https://decisions-api.net/api/v1/decisions \
  -H "Authorization: Bearer $DECISIONS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "decisions-1",
    "state": "The invoice total does not match the PO.",
    "questions": {
      "route": {
        "type": "choice",
        "instructions": "Which team should review this exception?",
        "criteria": {
          "ap": "Accounts payable or invoice mismatch.",
          "procurement": "Purchase order or vendor issue.",
          "fraud": "Suspected fraudulent document."
        }
      }
    }
  }'

FAQ